基于边缘计算的大规模无人协同探测感知架构研究

A Study on Large Unmanned Coordinated Detection and Sensing Architecture Based on Edge Computing

瞄准大规模无人集群等协同探测感知的高效应用,开展了基于边缘计算的大规模无人协同探测感知架构研究。以空海大规模无人雷达协同探测应用为例,介绍了该架构相比于云计算架构的优势;对信息传输量、数据延迟、传输能耗、体系抗干扰能力进行了定量分析与比较,并对边缘计算模型的目标检测损失进行了推导。研究结果表明:针对大规模无人协同探测感知应用,基于边缘计算的体系架构在较少的目标检测损失代价下,实现了信息传输量、数据延迟、传输能耗的大幅度降低,并有效提升了体系抗干扰能力,可鲁棒地支持大节点容量、高数据传输量、实时及资源受限的大规模无人应用场景。

To address the effective deployment of huge coordinated detection and sensing systems such as unmanned systems, coordinated detection and sensing architecture based on edge computing model is analyzed. Take the unmanned coordinated detection and sensing of targets on the air and sea as an example, the advantages of the architecture based on edge computing model are shown compared to that based on cloud computing model. The amount of transmission data, the data delay, the transmission energy, and the anti-jamming capability of the architecture are quantitatively analyzed in much detail. And the loss of target detection is analyzed as well. The research results show that for huge coordinated detection and sensing applications, the architecture based on edge computing provides less transmission data, data delay, transmission energy and more powerful anti-jamming capability at the cost of acceptable more loss of target detection. Therefore, it is suitable and robust for the real-time unmanned coordinated detection and sensing systems with great number of nodes, high amount of transmission data, and low energy storage.